The measurement methods covered by this part of GB/T6379 refer specifically to measurement methods that measure continuous quantities and take only one measurement value as the test result each time, although this value may be the result of a calculation of a set of observations. This part of GB/T6379 quantitatively defines the ability of a measurement method to give correct results (correctness) and the ability to repeat the same results (precision). This means that exactly the same thing can be measured using exactly the same method, and the measurement process is controlled. The main byproducts of GB/T6379 are applicable to a wide range of materials (substances or materials), including liquids, powders and solid materials. These materials can be artificially manufactured or naturally existing, as long as the heterogeneity of the materials is properly considered. GB/T 6379.1-2004 Accuracy (Trueness and Precision) of Measurement Methods and Results Part 1: General Principles and Definitions GB/T6379.1-2004 Standard Download Decompression Password: www.bzxz.net
The measurement methods covered by this part of GB/T6379 refer specifically to measurement methods that measure continuous quantities and take only one measurement value as the test result each time, although this value may be the result of a calculation of a set of observations. This part of GB/T6379 quantitatively defines the ability of a measurement method to give correct results (trueness) and the ability to repeat the same results (precision). This means that exactly the same thing can be measured using exactly the same method and the measurement process is controlled. The main byproducts of GB/T6379 are applicable to a wide range of materials (substances or materials), including liquids, powders and solid materials, which can be artificially manufactured or naturally existing, as long as the heterogeneity of the materials is properly considered.

1'5 3. 120. 30
National Standard of the People's Republic of China
GB/I ​​6379,1—2004/ISO 5725-1:1994 Partially recommended GT6379[86
GH/T:7921589
Aceuracy trueness and precision?of mcasurcmcnt mcthods and resaltsPart1:General principles and dcfinitions(IS0 5725-1:1994.IDT)
2004-06-02 Issued
General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China Standardization Administration of China
2005-01-01Implementation
Normative reference document
The actual meaning of the accuracy test definition
Standard test method
4,:
4.? 4.4 Experimental phenomena and laboratory observation conditions 4.5 Design models 4.6 Relationship between basic models and precision 4.7 Other possible models 4.8 Considerations in the design of tests for estimating accuracy 4.1 Principles of verification tests 4.3 Standard measurement methods 4.4 Experimental millimeters for accuracy tests Selection of test materials and the selection of accurate test materials - 7.1.1 Publication of accurate variation and precision values ​​7.2 Application of accurate and precise values ​​- Appendix A (Annex 4 of the Regulation) (B/T6HYD) Symbols and terminology used in Appendix B: Specification for uncertainty of precision variation points Appendix ((Informative) GB/T 6379.1-2004/[S 5725-1.1994 B/6379.1-2304/1605725-1.1994 GB/T37 The standard for the evaluation of the accuracy and precision of these methods and results is divided into six parts. The corresponding international standards are as follows: Part 1: General principles and definitions (ISO75-1:1991) Part 2: Basic principles for determining the repeatability and performance of standard force methods (ISO5725-2:1994. IDT) Part 3: Intermediate values ​​for standard measurement methods (corresponding to 1505725-3:1994) Part 4: This part of GB/T 637 is equivalent to the International Standard IS725-1:1994 Measurement methods and precision for special applications (Corresponding to TS725-5:1998) - Part 1: Alternative methods for determining the precision of standard measurement methods (Corresponding to TS725-5:1998) - Part 6: Accuracy of practical applications (Corresponding to IS) 5726-6:1991) - Part 1: General definitions and amendments to the 14th edition of GB/T 637. GI3/657S Part 1 to Part 6 replace GETe3791966GB/211792196S as a whole. The standard expands the original concept of precision, adds the concept of accuracy, and adds new provisions for repeatability and reproducibility. It also draws a section on precision: Appendix A of this part is a normative appendix, and Appendix C is a data collection appendix. This part is under the jurisdiction of the National Technical Committee for Standardization of Statistics. Drafting units of this part: China National Institute of Standardization, Institute of Mathematics and Systems Science of the Chinese Academy of Sciences, and Guangzhou Inspection Bureau of China. The main drafters of this part: Gan Ju, Ma Shi, Liu Wen, Qi Lian. Factory Wenxing, Xia Dou Wen, and Xiao Chu Li Cheng Institute. This part was issued on the 23rd time. GB/T63/9.1-2304/150 5/25-7:1994 Introduction
0.13.T637S uses the terms "correctness" and "precision" to describe the degree of accuracy that a measurement can achieve: The accuracy of a measurement is the degree to which the result is calculated. The reason for considering the density and viscosity range is that the same material is tested under the same conditions and generally gives comparable results. These three factors must be considered in order to ensure that the results are consistent: random variations will inevitably occur in each measurement sequence, and the results cannot be completely controlled. Therefore, process variation must be considered. If the difference between the test result and the measured value is within the range of the expected random error, in which case the test result must be consistent with the given value. The exact value between the values ​​does not need to be determined. For example, when comparing the results of two batches of material, the difference between them is not representative of the individual differences between the two batches. D.3 The results of the measurements made by different methods (except for the differences between the same samples) may vary, including: 1) changes in operators; 2) equipment used; 3) different equipment; 4) different temperatures; 5) different temperatures; 6) different humidity; 7) different humidity; 8) different humidity. Measurements made at different times and on different equipment will produce greater variation than measurements made by the same equipment over a period of time. The general term for density is pin density, and the two properties of pin density, namely repeatability and reproducibility, are sufficient for many actual flow shapes. For the measurement method, there is a certain degree of variability. Under the conditions of stress, all the above mentioned properties are not clearly defined and no variation occurs. Under small conditions, they are the starting point for the measurement: repeatability and reproducibility are the five conditions that provide the best quality for the product: stress provides the minimum variation, reproducibility provides the maximum variation. When one or more of the two conditions are met, the other conditions can also be found. They can be used for some specific circulations. The standard deviation is usually expressed as 0.5. When the true value of the property is missing or can be inferred, the accuracy of the measurement method is given by us. When it is recognized that the true value can be known with certainty. It is possible to know the "individual flow" of the test, for example, the appropriate standard can be used to determine the reference value of the test, and by comparing the pressure change value with the result level given by the quantity method, the correctness of the quantity method can be evaluated, and the proof can be confirmed by the use of the time and time again. Online shopping, the Eastern score is calculated, such as the use of the period or method to measure the total number of banned species, or to slide into a system and the quality of the system will determine the correct price of the cattle, 0.6[5>72 The technique used in the service is along the \correct" arrangement, which includes both the correctness and the accuracy of the system. This book Nolan used the cloud to express the time to survive. You want to make the previous score accurate, it seems to be enough for many people, the three-in-one result Liu transfer service (standard value system performance, should include the average of the expansion. For a long time, the term "price" has been used in statistics: out! It has aroused controversy in the medical and health circles, so the term "correctness" seems to emphasize its meaning. 1 Scope GB/T 6379.-2004/TSO 5725-:1994 Accuracy (trueness and precision) of measurement methods and results Part 1: General principles and definitions 1. ... Procedure for density measurement. Give its appropriate environmental benefit assessment II method (ISCI3725-): mountain) according to the basic force method (IS) 572 for the accuracy of the method of measurement accuracy for the specific application of the rate of different GB/5379.2 rent 150572-basic method within the assumptions of the visual measurement method to limit the accuracy and precision of other alternative forces (IS57253): to provide the correct safety and accuracy of the actual use of the work to improve the measurement method. 1.2 (3/T6779 this part of the sand covered by the measurement method.Special measures are taken for continuous and retrograde measurements, and only one required value is collected each time as the test result, although this value may be a set of constant expansion results. This part of T/T379 stipulates that the test method can give a unified result with the same accuracy as the test result. This means that the same method can be used to measure the same substance, and the test is scientifically controlled. This part of T/T8379 is applicable to a range of agents (substances or materials), including epidemic substances, powders and printed materials. These materials can be "manufactured" or naturally exist, as long as appropriate attention is paid to the heterogeneity of the materials. 2 Normative references
The following documents are used as references in this part of CB/T 370. The latest versions of the documents are not applicable to this part. However, the parties who have reached agreement on this part may consider whether to use these documents as the latest versions. The latest versions of the documents are applicable to this part.
G: B/T 3368.1-123 Statistical Specifications Part 1: General Statistics Terms GB/T 5358.2-1033 Statistics Card Language Part 2: Statistics - Quality Control Code G8/13358.3-1993 Legal Services Part 3: Verification Design Terms GDT 657$,2—200
Methods of measurement for accuracy of results (trueness and accuracy) Part 2, Standard methods for measuring the accuracy of dosage forms and the properties of small quantities of drugs S0341:1993 Statistics: 1 and symbols: Part 1, General statistical information S05, 4 Measurement methods and results (accuracy) Part 5, Standard methods for measuring the accuracy of results of drugs or small quantities of drugs SSC) 535-4:1S94 Measurement methods and standards for measuring the accuracy of dosage forms and the properties of small quantities of drugs (trueness and accuracy) Part 4, Standard methods for measuring the accuracy of dosage forms and the properties of small quantities of drugs S35-1:1993 Statistics: 1 and symbols: Part 1, General statistical information S05, 4 Measurement methods and results (accuracy and accuracy) Part 5, Standard methods for measuring the accuracy of dosage forms and the properties of small quantities of drugs SSC) 535-4:1S94 Measurement methods and standards for measuring the accuracy of dosage forms and the properties of small quantities of drugs (trueness and accuracy) Part 4, Standard methods for measuring the accuracy of dosage forms and the properties of small quantities of drugs S35-1:1993 Statistics: 1 and symbols: Part 1, General statistical information S05, 4 Measurement methods and results (accuracy and accuracy) Part 5, Standard methods for measuring the accuracy of dosage forms and the properties of small quantities of drugs SSC) 535-4:1S94 Measurement methods and standards for measuring the accuracy of dosage forms and the properties of small quantities of drugs S35-1:1993 5725-1:19943.1
Observed value (ohservedalue)
The value of a characteristic determined by a single observation. H9 K:T 3338, 1-134)
The value of a characteristic determined by a specified test method. (Revised K 338. —19:
Note 1: The test method should refer to the test card. The test results reported are the mean of the measured values ​​and other medical data such as the standard deviation of the measured values): it can be verified according to the applicable strict requirements. The gas is corrected according to the standard room and the force is corrected. Therefore, an example result can be calculated by a single test. In this case, the result is the test level of the least square test. 3.3
The level of the least square test 3.4
Cell in a precise experiment is the total average of all test results of a certain test material or sample. 3.5
Accepted reference value (accepted reference value) is a standard value for comparison, which is derived from: a) a theoretical value or a determined value established by a certain scientific institution; b) a specified value or a certified value based on the experimental design of some country or international organization. ：To a private school or group of people who have completed the whole teaching work of the school or group of people who have completed the certification: c
When a small amount of information can be obtained, the measurement period (type, egg specifies the value of the measurement summary 3.6
Accuracy (aocurac)
The degree of consistency between the test results and the accepted reference value (modified from GB/T35=8.11993)
Jiang; this language accuracy, when, the group test elimination mountain continuous error component and the family reading blue blood component are 3.7
Accuracy (lrueneas)
The average and numerical results obtained after a large number of measurements are consistent with the accepted reference value.
": The monitoring rate of the stop rate is small, the range of the month is small, this is called the "average efficiency", this usage is not distorted, 3.8
editor (HIA)
to the test result and the difference between the acceptance value. Note: On the contrary, the acceptance step is the sum of the system efficiency. The acceptable reason is: the human sense of the system error will cause the system error, the system error will eventually violate the resistance and be large, you or large,
laboratory seat bias laboratorblas
the test result of the special family is different from the acceptance requirement. 2
measurement bias (blas of 1hemeasirementmethad) CB/I6379.1—2002/NO 5725-1.1994 The difference between the expected result and the reference result obtained by the test method for all tests using the method. Juice 5: For example, when measuring the content of a compound in a control sample, the method has a drawback that many different laboratories using the same method may not be able to obtain the same result. Therefore, if the method used to measure the sample is different, the results may not be the same. 3.11
Biased test value (Luberatorstopomentnfhiag) The difference between the experimental value and the reference value: Note: The experimental value is more specific to the laboratory and may not be obtained under different test conditions. Student 8: The overall balance of the test results is related to the experiment, and There is no difference with the true value or standard value. 3.12
Precise density (precdsian)
Under the current conditions, the test results are established and the medicine should be used. (GT15R1993)
The quality is only related to the short-term distribution of the true value or the estimated value. Note: The precision of the information is not selected by the market. The quality of the test results is not shown. The material is brought to the description by the standard. Note 11, "maintaining the results" means that the results of the test are not affected by the previous results. The conditions are strictly low. The multiple mixing of the fixed basis. The two conditions are repeated and the quality is repeated. 3.13
Repeatability (repeatabiit)
Under the left side of the precision of the design
(change 338, 1—
Set the repeatability conditions (rpewtubilitycorklitions) in the same laboratory, by the same, the same equipment, the same test method, in the knowledge of the internal and external test objects, the test results are mutually trusted,
(GT/T 33.8. 1—19U)
The repeatability standardderlation is the standard for the test results compiled under the repeatability conditions. (E from GB33=6.1S93)
: You are here
Calculation 1]: The repeatability coefficient is used as the factor of the repeatability coefficient before the test, 3,15
Repealubililyliunit is two, a value, under the repeatability conditions, the absolute difference between the two initial stations is less than or equal to the inverse number with a probability of 95%. (GBT335&.11S93)
Design 1-, manual reproducibility is shown in a book,
Reproducibility (reprdueibility)
The density changes under reproducibility conditions.
(Adopted from GB/T 33.1--1593)
GB/T 6379.12004/1S0 5725-1; 19943.1e
Reproducibility conditions (iepraducihilityeunditionx) are the evaluation of the test results of the same test object under the same test method in different laboratories without different operators using different poles.
(Adopted from GB/T 3358.1--1593)
GB/T 6379.12004/1S0 5725-1; 19943.1e
Reproducibility conditions (iepraducihilityeunditionx) are the evaluation of the test results of the same test object under the same test method in different laboratories without different operators using different poles.
(Adopted from GB/T 3358.1--1993)
Reprodribilltystandarddeviation is the standard for the test results under the same test conditions. -—199J)
Note 15: The reproducibility test will be repeated in the next test. The reproducibility test results can be determined by comparing the results of the previous test with the current variability. The dispersion ratio is: 3.2
Reproducibility limit (reprontucibilitylimit): number, under the condition of actual phenomenon, the absolute difference of the test results is less than or equal to this number, (G3/T 53=8.11393)
17: The actual limit is equal to the scale
commercial group estimate (outlicr)
one or more test values, which are far away from other actual values, indicating that there may be some unreliable totals, (from G8,19)
reproducibilitylimit: GB/6379.2 meets the requirements of positive variability and precision variability. Sichuan comes to plan commercial sample risk inspection and significance level, 3.22
callaborativeassessmentexperlment: teacher Laboratory planned clinical tests, in this test will be, with the same material tested using the same standard method or formula to assess the technical criteria of each experimental case
juice [9: in 3.1, 2 given mountain of the specified text using the "observation" principle, the results of the test or is the matter of glue or after rapid agreement, then the above is the significance of the minimum conditions of the respective methods: the absolute value of the test is less than or equal to the yield of the tooth value is not less than %.
Note: mountain 3. to 11, .15, 31%, 4.19 and %.23 in the text of the Ministry of protection light screen application this 3 theory, the actual determination of the amount of Xia Dan to the ring difference and the transmission of the physical - Chuan 62 and 2 in the language of the letter of the letter of the letter of the letter, because the ratio will be beneficial,,,,: the water half will not be good, etc., very practical room! , this is a close to the actual number of shares with the actual number of ten, the level may need to be 5 relatively selected, this is not the perfect world, do not underestimate its actual effect, because the original design of it is only as a one-shot, to judge whether the actual station is formed by the uncertainty of the base of the figure. The North Heavy Method should be re-verified to cause a large difference in the complete note, note 1: the symbol, and R also have a more The meaning of the IS9+-1 is that the correlation number, RW, is expressed in terms of the value of the bamboo, which is sufficient to produce, and the reference number is the position of the weighing complex, or the limit of the formation, so that the sample will not fall into the courtyard.
4 The actual meaning of the definition of the wall test
4.1 Standard measurement method
4.T.1 The fire standard measurement is carried out according to the same point method, and the measurement method should be standardized. All parts should be selected according to the specified standard method. This means that there must be a written document that stipulates all the requirements of the plan, and also includes how to obtain and prepare the test column:
GB/T 6379.12904/ISO 5/25-1,19944.1.2 The production of a document on measurement methods implies the existence of an organization responsible for the research and development of measurement methods. 4.2 Accuracy test || tt || 4.2.1 Accuracy (correctness and precision) The test is for the laboratory to determine the accuracy and precision of the test results. There is no special expert group organized to organize all the tests. This test is not designed for the laboratory and is called an "accuracy test". It can also be called a "test" according to its limited test limits. Density test or "on", if the standard is determined before, then the accuracy value obtained by the test should be indicated, and the measurement method used should be specified. The result is only valid under the method used.
4.2.2 Accuracy or verification method can usually be considered as a standard test. The accuracy of a standard can usually be used to estimate the suitability of a product (experimental) for a certain period of time. The evidence provided by the accuracy test will show how valid the standard is. The within-test variances seen in Chapter 1 or the differences between tests may indicate that the standard measurement method is not detailed enough and can be improved. For example, the problem may be referred to a standardization body for further improvement. 4.3 A test plan
4.3.1 In a standardized test, samples of a product are sent from a central point to a number of laboratories located in different locations, in different countries, or even in different continents. Definition of the repeatability condition (3.14) The tests carried out in these laboratories may be carried out on the same test objects and in the same period of time: to achieve this the following two conditions must be met: b) the samples should be identical in time spent in transport and in the test before being measured. In the test, care must be taken to ensure that this condition is met. 23: 0.1 More discussion of the method of linking materials, 4.4 Short time intervals || tt || 4.4.1 Three conditions for repeatability (5.14). Repeatability tests must be carried out under the same conditions; that is, the factors listed in .3 must remain constant throughout the test period. No recalibration should be done between measurements. Individual measurements or parts of the same test should be made. In practice, tests should be carried out under repeatability conditions for as long as possible to minimize factors that cannot always be verified, such as environmental changes. 4.4.2 Another factor that may affect the test results is the assumption that the time between different tests is the same. To avoid the possibility that the test results may be affected by the following test methods: It is necessary to provide samples in the following way: Operators should follow the same sample. The operator follows a certain routine, but the parameters are random, so that all tests of the "same" object are not performed together, which may actually violate the requirement of repeated measurements within a short period of time: unless a new test mouse can be completed within a short time. 4.5 Laboratory
4.5.1GB/6! Some of the basic levels of the standard are assumed to be benchmark measurements. Of course, each laboratory in this standard library should be or at least approximately the same. This allows for an average standard deviation of its results, which is applicable to any laboratory. However, some laboratories should make a series of observations under the same conditions. Each laboratory should make its own estimate of the standard deviation for the measurement method and calibrate the estimate according to the standard deviation. 1S) 37-6 discusses this approach in detail,
4.5.2 In the definition of 3.8 to 3.7:1, in theory: E applies to all experiments that might use the test method, in practice - they are the same as the ones that were originally determined by the laboratory. This will be discussed in detail in the next section. When the test method is extended to the standard in E.3, the required accuracy and risk of the test can be estimated to meet the product requirements. However, if there is evidence that the experiment involved in the test is not or no longer truly representative of all laboratories that test the method, the new method must be carried out. 4.6 Observation conditions 4.6.1 In 0.3, all factors that may affect the results of the observations obtained in one laboratory shall be noted, including time, operator and equipment. Because when the tests are performed at different times, changes in the environment, changes in equipment control standards, etc. will affect the values ​​obtained under the same conditions, the values ​​obtained in different laboratories after observation, all other factors will change, and the management and maintenance between the two laboratories and the stability of the observed values ​​will also have an impact on the results. 4, 6, 7 Sometimes there are intermediate density conditions that must be checked, that is, the observed values ​​are obtained in the same laboratory, but one of the factors, time, sample preparation, or equipment has changed. In determining the consistency of the test method, it is more important to observe the observation conditions, that is, the time, working point, or equipment mentioned above - these factors remain unchanged, and none of them changes. In addition, the difference caused by this factor is related to the whole process. For example, in chemical analysis, "operator" and "time" are important factors; in micro-analysis, "preparation" and "environment" play an important role. In physical testing, "equipment" and "calibration" are the main factors.
5 Statistical model
5.1 This model
is used to calculate the accuracy (precision and accuracy) of the separation method. For a given critical material, each test result is a sum of the values ​​of the laboratory's total mean value (bottleneck) under reproducible conditions: (a) the random error of the oxygen production per source under reproducible conditions. 5.1.1 Total mean value m
5.1.1.1 The half-mean value is the test level, and different samples of a chemical component (e.g. different types of samples) correspond to different levels. In many cases, the test level is determined only by the test method and the concept of true value is not used. However, in some cases, the concept of true value for the test component can still be used: for example, the true concentration of a titrated solution. The half-mean value may not be the same as the half-mean value. 5.1.1.2 The half-mean value is not necessarily the same as the half-mean value. The half-mean value is not necessarily the same as the half-mean value. However, when comparing test results with values ​​specified in a standard or in a test solution, where the standard or the test solution is not true, or when comparing results obtained by different methods, the bias of the measurement method must be considered. If there is a value for which a good reference can be obtained, then the method of JS5725-4 should be used to determine the value of the test method.
5.1.2 Analysis B
5.1.2.1 In any series of tests conducted under mutually exclusive conditions, analysis B may be considered normal, but tests conducted under other conditions may not be normal. When comparing two identical test chambers, it is necessary to determine their corresponding corrections: by determining the bias of the test chambers themselves through a standard test, or by discarding them from a special test set. However, when comparing two identical test chambers, or when comparing two test chambers that do not share any common characteristics, it is necessary to consider the difference in the experimental bias component. The variance given in GB/373 is obtained under the assumption that the distribution is unimodal. In most practical cases, it is sufficient to assume that the distribution is unimodal: 5.1.2.2 The variance of the experimental room is expressed as follows: vari(B) = a1
where it includes the variation between operators and equipment, ...r.
In the basic precision test of GB/637., these components are slightly decomposed: The method of giving the size of the random component of the distribution is given in IS() = 25-3. 5.1.2.3 Abnormally, B can be the sum of the random component and the system component. This does not include all the factors that are related to the system. CB/T 6379.1-2004/IsU 5725-1,1994 These factors include different climatic characteristics, the manufacturer's allowable equipment variation, and the technical differences caused by the operator's training at different locations. 5.1,3 Error items 5.1.3.1 The error produced by each test result will be shown in the error table. In this part of H637: All the difference sequence errors are approximated by normal distribution. In most practical cases, it is only necessary to assume that the error is a single choice. 5.1.3.2 The variance of a single laboratory under five medical conditions is called the experimental internal variance, which is expressed by the following formula: var(e) =
5.1.3.3 Due to factors such as the selection technique, the value of the test child may not be changed: However, it can be determined by general standardized measurement methods. The difference between laboratories is not small. For all laboratories using the pilot method, an indoor variance that is equivalent to that of each laboratory can be set. This variance is called the experimental variance and can be calculated by calculating the intra-laboratory variance: As shown below, vur(e)
The arithmetic mean in the above formula is the actual integration period after eliminating outliers: 5.2 Relationship between basic model and precision
5.2.1 When the basic model in 5.1 is used, the variance of the required error can be directly used as the error term. However, the variance is the sum of the variance of the experimental children and the variance between the experimental children in 5.1.2.2. 5.2.2 As two quantities of the product, the standard deviation of the reproducibility
c, = Varr)
Reproducibility
R-Va+eat
5.3 Other alternative models
There are 6 alternatives to the product model that need to be considered, which are fully described in the other relevant parts of G/15379. 6.1 Planning of an Accuracy Test
6.1.1 The design of an accuracy test should include a team of experts who are familiar with the measurement method and the task at hand. At least one of the experts should have experience in statistics and experimental analysis. 6.1.2 When planning an experiment, consider the following questions: Is there a satisfactory standard for the measurement method? 5) How many laboratories should be involved in the experiment? Which laboratory should meet the requirements?
|| How many levels should be repeated in the experiment: What materials are used to express these levels? How many times should the materials be repeated?
|| How long should all these plans be completed?
|| 5. Is the original model suitable for it? Does it need to be modified?
|| 6.1.2 When planning an experiment, consider the following questions:
|| How long should all these plans be completed?
|| 6.1.3 Is the original model suitable for it? Does it need to be modified?
|| 7. ) Are special precautions required to ensure that the same material is measured under the same conditions in different experimental conditions? These issues can be considered in Section 6.2. Standard measurement methods
As pointed out in Section 4.1, the method used for the new test method should be a standardized method. Such a method should refer to GR/1 6379.1-2001/1S0 5/25-1:1994, that is, the measurement result will not produce unexpected large changes due to small changes in the measurement process. There should be appropriate precautions or warnings. In the preparation of a standard measurement method, every effort should be made to minimize the degree of reliance on the method. It is also possible to use some kind of test procedure to improve the accuracy and precision of the new standard measurement method. In the latter case, the values ​​obtained are considered preliminary and are considered to be accurate as the accuracy changes with laboratory experience. The documentation of the measurement method should be complete. All essential operating procedures involving the environment, reagents and equipment, initial testing and preparation of the test specimens should be included in the measurement method. These methods should also be described in a document that is as informative as possible to the operator. The instructions should state the test results and calculations precisely and the number of valid digits that should be reported. 6.3 Choice of test room for accuracy 6.3.1 Choice of test room for accuracy From a statistical point of view, those experimental rooms in which the accuracy of the measurement method is to be estimated should be compared with those in which the measurement method is to be performed, as experimental conditions that are not optimal for the laboratory may not be adequate for the laboratory. Of course, one of his concerns is that the number of tests to be carried out may have an impact on the performance of the test, for example by requiring that the tests be carried out in different continents or different geographical locations. The participating laboratories should not consist only of those which have acquired special experience in standardizing the measurement of power, and should not be composed primarily of those special "standard" laboratories which maintain that the "standard" laboratories are experts in demonstrating that the test is indeed accurate.
It is also important to determine the number of tests to be carried out in the collaborative study and the number of test results to be carried out at each test level. Guidelines for determining these numbers are given in 5.3.3 to 6.3.4. 6.3.2 The precision required for the test is bzxz.net
6.3.2.1 In the first investigation, the symbol 2 represents the unknown standard deviation. Then the precision test is estimated: When the estimated value can be obtained for the true value of the standard, the relevant general theory can be obtained, that is, the estimated value: the expected value in the norm, which is a well-known statistical problem, can be obtained by the distribution and: the estimated value is based on the test year and the test is effective. The formula used in the general book is:
P: :A2P4+A) =H
A represents the cumulative effect of the uncertainty of the estimated value, which is usually expressed as a percentage. (7) Formula indicates that the estimated value of the standard deviation can be expected with probability P (4 times the difference between the two tests). 3.2.2 Single country production, the actual standard deviation of the small standard deviation according to the norm and the actual case of the whole case error to the number. For the reproducibility standard deviation, a more accurate calculation procedure is required (Equation (6)). In this case, another factor is required to express the ratio of the reproducibility standard deviation to the reproducibility standard deviation; 7 = GR/D,
6.3.23 The following approximate formula is used to calculate the number of laboratories required for the reproducibility level 5 and the number of test results required for each laboratory at each test level: These equations have a fixed limit, so the limit calculation method is not used. The approximate formula for reproducibility A is as follows:
for reproducibility.
A-A1.96,/21+-(-1)
2y*n(p?
4It can be generally determined that there is an approximate formula (the formula is the inverse of the law of ratio, which can be verified by the above approximate formula through frequency counting.
Tip: This standard content only shows part of the intercepted content of the complete standard. If you need the complete standard, please go to the top to download the complete standard document for free.